And in the light of Swift's latest researches in the case of the sex cords and the ova of the chick ovary and the spermatozoa of the testis, it now seems clear (notwithstanding the contradictory findings of Dustin in Chrysemp and of Gatenby in the frog) that the definitive germ-cells have had an extra-regional origin and do not arise by differentiation from germinal epithelium (ccelomic epithelium), but become involved in columnar invaginations of these cells to form sex cords. Certain observations (e. g., those of Felix, Dustin, and Firket) to the effect that the primordial germ-cells (" primary genital cells") degenerate early and are replaced by ccelomic derivatives in the genital gland, may possibly be explained by the facts that in certain amniotic forms the primordial germ-cells degenerate relatively more extensively and that mitosis does not generally appear until the remainder reach the sex-glands, where it becomes active in those cells involved in the sex cords formed by the "germinal epithelium." The great proliferative activity of the remaining primordial germ-cells in the formation of oiigo nia and the spermogonia, among the cells of the ccelomic epithelium, gives the appearance of an active differentiation and derivation from the ccelomic epithelial-cells. The products of this intense proliferation are relatively smaller than the earlier germ-cell progenitors, which phenom enon makes the superficial resemblance between growing ccelomic epi thelial cells and the definitive germ-cells more close. At any rate, no clear case has yet been made out for the differentiation of germ-cells from ccelomic epithelium at any stage in any form.
There is no evidence in Caretta of a differential mitochondrial content between the germ-cells and somatic cells, as maintained by Tschaschin in the case of the chick embryo and by Rubaschkin in the case of certain mammalian embryos; and as denied by Swift and von Berenberg-Gossler. The germ-cells are distinguished from the so matic cells by their larger size, their generous yolk-content, their large vesicular oxyphilic nucleus, with its delicate, finely granular, radially arranged, nuclear network. These nuclei resemble more closely those of the entoderm-cells of the area opaca, indicating a low grade of differentiation of these two cells and a close genetic relationship inhering most probably in a common near ancestor, from which they have departed but little in their slight progressive differentiation. This indicates also that the germ-cell commonly remains undifferentiated until a relatively late period in its history, when it takes residence in the genital gland. Its load of nutritive material in the form of yolk (still present in the 25-day stage) also indicates a low grade of differen tiation and suggests the cause of its lack of proliferative capacity until digestion of this yolk is completed, and the further reason of its great capacity for later proliferation and differentiation, especially in the male.
But many more researches covering the later periods of the history of the ovary with special reference to the so-called "germinal epithe lium," such as those of Swift on the chick and of Gatenby on the frog, are needed before the hypothesis of a germinal path for vertebrates can be said to be completely established. In view of conditions in the
gonads of certain annelids, mollusks, and echinoderms, where enormous numbers of germ-cells are formed during successive years, and the disagreements which still exist with respect to such relatively simple forms as frog (Allen and King vs. Dustin and Gatenby), chick (Swift vs. Firket), and the turtle, Chryaemys (Allen vs. Dustin), the claims that two series of genital cells occur—a primordial and a secondary or definitive—the one derived directly from a blastomere of the later cleavage stages and so segregated from the somatic cells, and the other derived by differentiation from the ccelomic epithelium of the genital ridge, and so modified soma-cells, while on logical grounds inherently improbable, and without firm observational basis—and disproved on histologic grounds, I believe, in the case of Caretta—can not perhaps be said to have been definitely disposed of.
Summary OF RESULTS.
1. Twenty-five embryos of the loggerhead turtle (Caretta caretta), ranging from the second day (5 somites, 2 mm. length) to the thirty second day of incubation, were employed in this investigation. The early history of the primordial germ-cells is very similar to that de scribed by Allen for the turtle Chrysemys and by Woods for the dog fish.
2. The primordial germ-cells migrate during the second day from the yolk-sac entoderm, where they were more or less widely scattered caudally, into the lateral border of the area pellucida on each side of the embryonic disk. Here they become sharply segregated by the beginning of the third day into two bilateral cords situated in the entoderm of the area pellucida laterally in the caudal half of the disk. In the 2-day embryo they extend from the neurenteric canal to the end of the primitive streak; in the 3-day embryo from the sixth somite to the caudal extremity of the streak. The cords become more medially placed, make a linear connection with the overlying visceral mesoderm, and their cells migrate during the fifth day into this mesoderm, and thence medially (during the sixth and seventh days) towards the root of the forming mesentery of the closing hind-gut. Individual cells migrate medially also within, or back into, the entoderm of the gut. The germ-cells in the medial entoderm become included in the mucosa of the closed hind-gut, those in the mesoderm in the enveloping mesenchyma and the gut end of the mesentery. From these locations the majority of the germ-cells subsequently (seventh to twelfth day) migrate up the mesentery and across the ccelomic angle into the future sexual gland. They become incorporated among the mesenchymal cells of the gland and the covering peritoneal epithelium, where they suffer no striking change in form, size, or content at least as late as the thirty-second day of incubation.